A Revision of the 1783?1784 Calabrian (Southern Italy) Tsunamis L

A revision of the 1783?1784 Calabrian (southern Italy) tsunamis L. Graziani, A. Maramai, S. Tinti

To cite this version:

L. Graziani, A. Maramai, S. Tinti. A revision of the 1783?1784 Calabrian (southern Italy) tsunamis. Natural Hazards and Earth System Sciences, Copernicus Publ. / European Geosciences Union, 2006, 6 (6), pp.1053-1060. hal-00299395

HAL Id: hal-00299395 https://hal.archives-ouvertes.fr/hal-00299395 Submitted on 13 Dec 2006

HAL is a multi-disciplinary open access L’archive ouverte pluridisciplinaire HAL, est archive for the deposit and dissemination of sci- destinée au dépôt et à la diffusion de documents entific research documents, whether they are pub- scientifiques de niveau recherche, publiés ou non, lished or not. The documents may come from émanant des établissements d’enseignement et de teaching and research institutions in France or recherche français ou étrangers, des laboratoires abroad, or from public or private research centers. publics ou privés. Nat. Hazards Earth Syst. Sci., 6, 1053–1060, 2006 www.nat-hazards-earth-syst-sci.net/6/1053/2006/ Natural Hazards © Author(s) 2006. This work is licensed and Earth under a Creative Commons License. System Sciences

A revision of the 1783–1784 Calabrian (southern Italy) tsunamis

L. Graziani1, A. Maramai1, and S. Tinti2 1Istituto Nazionale di Geoﬁsica e Vulcanologia, Sezione Roma 2, Rome, Italy 2University of Bologna, Dipartimento di Fisica, Settore Geoﬁsica, Bologna, Italy Received: 28 September 2006 – Revised: 23 November 2006 – Accepted: 23 November 2006 – Published: 13 December 2006

Abstract. Southern Italy is one of the most tsunamigenic 1 Introduction areas in the Mediterranean basin, having experienced during centuries a large number of tsunamis, some of which very In the Mediterranean basin, Italy is one of the countries most destructive. In particular, the most exposed zone here is the affected by tsunamis. During centuries it has experienced a Messina Straits separating the coasts of Calabria and Sicily large number of events, some of which caused catastrophic that was the theatre of the strongest Italian events. In 1783– effects, severe damage and thousands of victims. The analy- 1785 Calabria was shaken by the most violent and persis- sis of the Italian Tsunami Catalogue (ITC, Tinti et al., 2004) tent seismic crisis occurred in the last 2000 years. Five very shows that southern Calabria, Messina Straits and eastern strong earthquakes, followed by tsunamis, occurred in a short Sicily are certainly the most active tsunamigenic regions in interval of time (February–March 1783), causing destruction Italy. As a matter of fact, the majority of the Italian tsunamis, and a lot of victims in a vast region embracing the whole including the strongest ones, occurred here: most were the southern Calabria and the Messina area, Sicily. In this study results of coastal and submarine earthquakes, but some were we re-examined these events by taking into account all avail- also due to volcanic activity and submarine slides. able historical sources. In particular, we focussed on the 5 The geological setting of the area is quite complex and and 6 February 1783 tsunamis, that were the most destruc- some aspects are still poorly understood. Structural anal- tive. As regards the 5 February event, we found that it was ysis and seismological and deformation (VLBI and GPS) underestimated and erroneously considered a minor event. data show a regional extensional ESE-WNW trend which is On the contrary, the analysis of the sources revealed that in coupled with a strong 1.5–2 mm/yr uplift of the whole Cal- some localities the tsunami effects were quite strong. The 6 abrian arc region (Monaco et al., 1996; Tortorici et al., 2003; February tsunami, the strongest one of the sequence, was due D’Agostino and Selvaggi, 2004; Jenny et al., 2006). One of to a huge earthquake-induced rockfall and killed more than the main tectonic features is the belt of normal faults running 1500 people in the Calabrian village of Scilla. For this event for more than 350 km along the Apennines chain and extend- the inundated area and the runup values distribution were es- ing from Calabria to Sicily, which is formed by distinct fault timated. segments with maximum length of about 40–50 km that are Further, the analysis of the historical sources allowed us responsible for the major earthquakes. Some of the faults are to ﬁnd three new tsunamis that passed previously unnoticed found close to the coast or are completely placed offshore and that occurred during this seismic period. The ﬁrst one with great potential for tsunami generation. The most ener- occurred a few hours before the large earthquake of 5 Febru- getic historical shocks reached the maximum magnitude of ary 1783. The second was generated by a rockfall on 24 M∼7 (Boschi et al., 2000) and almost all were accompanied March 1783. Finally, the third occurred on 9 January 1784, by sizable tsunamis. probably due to a submarine earthquake. Before focusing on the 1783–1785 seismic period, it is worth mentioning the two largest tsunamigenic earthquakes that occurred in Calabria and Sicily, namely the 1693 eastern Sicily and the December 1908 Messina Straits events (Fig. 1). Correspondence to: A. Maramai At the beginning of 1693 a tremendous seismic crisis hit ([email protected]) eastern Sicily, culminating in the main shock of 11 January

Published by Copernicus GmbH on behalf of the European Geosciences Union. 1054 L. Graziani et al.: Revision of the 1783 Calabrian tsunamis

39.200 N sequence was exceptionally long and destructive, the associ- 14.300 E CALABRIA ated tsunamis were almost disregarded as secondary minor N TYRRHENIAN events, except for the 6 February Scilla tsunami that caused SEA much more casualties than the parent earthquake itself. This unusual and long seismic period stirred up great interest in many Italian and foreign scientists and scholars, that went on the field to make direct observations and to document the effects. As a result, a vast amount of reports, chronicles, 1908 works and maps representing the devastation was produced Messina Scilla and this constitutes today a precious source of information Reggio Calabria for studying the earthquake effects and consequences, in- Pellaro IONIAN cluding tsunamis. By carrying out a deep analysis of the S. Alessio SEA available sources it was possible to count as many as ten dis- Giardini Naxos tinct tsunami events, most of low intensity, three of which SICILY were never mentioned before in the modern scientific litera- Catania ture. For the largest tsunamis, the available data allowed us to draw a rather reliable picture of the effects and, in particu- 1693 Augusta lar, for the 6 February Scilla event it was possible to perform the reconstruction of the inundated area and to determine the Siracusa distribution of the run-up values. This work contributes to the updating of the ITC (Tinti et Noto 36.800 N 17.400 E al., 2004) and, for this purpose, the events studied here were classified according to the evaluation standards adopted for Fig. 1. Epicentres of the earthquakes generating the most destruc- the catalogue compilation. tive tsunamis in southern Calabria and eastern Sicily. Localities most affected by waves are reported. 2 The 1783–1785 Calabrian seismic sequence with estimated magnitude M=7.4 (CPTI2, 2004). It pro- In 1783–1785 the southern part of Calabria was struck by duced almost complete destruction within an area of about one of the most impressive and persistent seismic crisis that 2 14 000 km extending from Catania to Noto and Siracusa, occurred in the last 2000 years. A lot of sources are avail- causing more than 54 000 casualties (Boschi et al., 2000). able in the literature: many famous coeval writers and a A tsunami was observed along the whole eastern Sicily coast lot of current historiographer studied and documented this with major effects at Augusta and Catania. First the sea with- seismic period. During our study we examined more than drew from the shoreline and the following inundation was 100 bibliographical sources, the most reliable ones, and we very damaging and lethal for many people. At Augusta, the also performed extensive research in many public and private water waves were likely as high as 15 m (Tinti et al., 2004). archives, libraries, etc., in order to collect new data. The cri- The 28 December 1908 Messina earthquake is one of sis lasted more than three years, starting at the beginning of the strongest earthquakes that ever occurred in Italy. The February 1783 and, during a period of about two months five towns of Messina and Reggio Calabria together with many large earthquakes (with epicentral intensity MCS≥IX) oc- other villages were completely destroyed. Partial destruc- curred along a zone about 100 km long, devastating southern tion involved most of Calabria and Sicily over an area about Calabria between Reggio Calabria and Catanzaro. The first 6000 km2 wide. More than 60 000 people died. A violent destructive earthquake of the sequence ocurred on 5 Febru- tsunami followed the shock, causing severe damage and hun- ary 1783, and was followed by the other main shocks re- dreds of victims in the Messina Straits zone. In most places spectively on 6 and 7 February, and on 1 and 28 March in the sea first receded for a few minutes from the usual level the same year. All these earthquakes resulted to be tsunami- and then flooded the coast with at least three big waves. The genic, though the size of the generated tsunamis differed very sea level oscillations lasted for many hours. The tsunami much from case to case. The epicentres of the five main reached its maximum intensity in the Calabrian coast near shocks migrated northward from the Aspromonte area to the Pellaro, where run-up heights as large as 13 m were observed, Sant’Eufemia and the Squillace gulfs (Fig. 2). In addition and in Sicily at Giardini Naxos and at Sant’Alessio, where to these main earthquakes, hundreds of smaller aftershocks the measured run up was 11.7 m (Tinti et al., 2004). hit southern Calabria. The seismic sequence was one of The main purpose of this paper is to provide a picture of most catastrophic in western Europe. The cumulative effects the tsunamis that occurred during the seismic crisis that dev- of all these earthquakes was devastating and great changes astated the Calabria region in 1783–1785. Since this seismic were produced even in the environment morphology: diffuse

Nat. Hazards Earth Syst. Sci., 6, 1053–1060, 2006 www.nat-hazards-earth-syst-sci.net/6/1053/2006/ L. Graziani et al.: Revision of the 1783 Calabrian tsunamis 1055

39.200 N 14.300 E Table 1. Reliability scale of the Italian Tsunami Catalogue (Tinti et N CALABRIA al., 2004).

S. EUFEMIA GULF Catanzaro 01/03/1783 Reliability scale of the Italian Tsunami Catalogue 28/03/1783 Tropea SQUILLACE 0: Very Improbable Tsunami GULF 07/02/1783 1: Improbable Tsunami TYRRHENIAN SEA Gioia Tauro Stilo 2: Questionable Tsunami 3: Probable Tsunami Bagnara 05/02/1783 IONIAN SEA 06/02/1783 4: Deﬁnite Tsunami Messina Scilla ASPROMONTE Reggio Calabria SICILY

37.800 N On 7 February another big shock, with epicentre located 17.400 E about 40 km north-east to the 5 February event, hit southern Fig. 2. Epicentres of the five main shocks of the 1783 Calabrian Calabria. The shock (Mw=6.6, I=X–XI MCS; CPTI2, 2004) seismic crisis. Square size increases with magnitude. occurred at 13:10 and caused new heavy destruction. On 1 March at 01:40 the fourth big shock occurred. The estimated magnitude was Mw=5.9 (CPTI2, 2004) with I=IX MCS and the epicentral area was located about 20 km north landsliding led to heavy changes to the hydrogeological sys- of the 7 February earthquake, along the Apennines chain. tem (more than 200 lakes formed as a consequence of stream The last big earthquake of the sequence occurred on 28 damming caused by slides). The most extraordinary ef- March at 18:55. The epicentre was slightly west to the fourth, fects were chiefly due to the 5 February earthquake (Viven- and the estimated magnitude was Mw=6.9 with I=X MCS zio, 1788; Hamilton, 1783; Dolomieu, 1784; Baratta, 1901; (CPTI2, 2004). Jacques et al., 2001). On the 5 February, at about 12:00, a catastrophic tsunami- 3 The 1783–1784 tsunami events genic earthquake abruptly opened the seismic period. The shock (Mw=6.9, I=XI MCS; CPTI2, 2004) was felt in a very In order to perform the reconstruction of the tsunami events wide area, including the whole Sicily and a big portion of that occurred during the 1783–1785 seismic crisis, we gath- the southern Italian peninsula. More than 380 villages were ered the historical material from contemporary chronicles damaged, 180 were almost totally ruined and the count of fa- and reports available in national and local libraries, private talities exceeded 25 000. Major effects were observed close and ecclesiastic archives, etc. The analysis led us to an im- to the Gioia Tauro plain (on the Tyrrhenian side), at the west- proved knowledge of the main events already present in the ern foot of the northern Aspromonte, where some towns were ITC and to discover three new tsunami events. The tsunamis completely destroyed and many other suffered severe dam- were examined following the ITC guideline: in particular age. The Ionian side was less affected. Reports tell that at the each event is characterised by a tsunami intensity value, a time of the shock people in the fields were thrown down and reliability score and, when known, by the generating mecha- trees were uprooted. Though this event has been widely stud- nism. As far as the intensity is concerned, the 6-degree inten- ied, the identification of the generating fault is still debated. sity scale proposed by Sieberg and Ambraseys (Ambraseys, Tsunami information can be used to better define the possi- 1962) is used to quantify the effects of a tsunami on struc- ble source (Tinti and Piatanesi, 1996a), as was also shown for tures and environment. An intensity 1 tsunami is “very light” other cases, where comparison between the available tsunami and only visible on tide gauge records, while at the top of the data and tsunami simulations was used to constrain the focal scale an intensity 6 tsunami is capable of causing large disas- mechanism as well as to exclude some of the fault locations ters with total destruction of man-made structures, extensive proposed in literature (see e.g. Piatanesi et al., 1996; Tinti flooding, many casualties, etc. Notice that we also attributed and Piatanesi, 1996b). the tsunami intensity on the basis of the 12-degree scale that The shock of 6 February occurred during the night (00:20), was proposed by Papadopoulos and Imamura in the course of was less strong than the previous day earthquake (equiva- the 2001 International Tsunami Conference, but that was not lent magnitude from macroseismic data Me=6.3, I=VIII–IX; adopted in the ITC first version (Tinti and Maramai, 1996) CPTI2, 2004) and was located slightly to south-west, in the nor in its recent update (Tinti et al., 2004). Messina Straits. It caused severe damage at Messina and at The tsunami reliability is graded according to the 5-degree Scilla. The earthquake was responsible for a huge rockfall modified Iida scale and ranges from degree 0, equivalent to that occurred along the western cliff of the Mount Campalla` a “very improbable tsunami”, to degree 4 corresponding to at Scilla, that fell into the sea generating a disastrous tsunami. a “definite tsunami” (see Table 1). The reliability is a very www.nat-hazards-earth-syst-sci.net/6/1053/2006/ Nat. Hazards Earth Syst. Sci., 6, 1053–1060, 2006 1056 L. Graziani et al.: Revision of the 1783 Calabrian tsunamis

39.200 N 38.800 N 15.200 E CALABRIA 14.800 E CALABRIA N N Cutro Capo Vaticano Joppolo S. EUFEMIA GULF Nicotera Capo Rizzuto Gioia del Tirreno Gioia Tauro Pizzo Calabro SQUILLACE GULF Bivona Torre del Faro Roccella Ionica Scilla Messina Cenidio Catona Reggio Calabria Bianco

IONIAN SEA 37.600 N 17.000 N 37.900 N 17.400 E Fig. 4. Epicentre of the 5 February 1783 main shock. Localities hit Fig. 3. Localities involved by the tsunami occurred in the morning by the generated tsunami are reported. of 5 February 1783, at 08:00.

to its usual limit (Vivenzio, 1788). On the other hand, De important parameter which helps to judge the quality of the Lorenzo (1877) reports that at Bivona and at Pizzo Calabro, data. The approach for determining reliability value con- in the Tyrrhenian Calabria, some fishermen on the boat noted sists of taking into account and rating separately different that in the morning the sea became suddenly agitated. The elements or categories of data, namely: 1) what is known only seismic catalogue reporting an earthquake in the morn- about the tsunami cause, 2) what is known about the tsunami ing is the Istituto Nazionale di Geofisica catalogue (ING, itself, and 3) what is known about the documentary sources. 1991), in which a weak shock.(I=III MCS scale) with epicen- Each category is rated on three levels, and then the final relia- ter located at capo Rizzuto on the Ionian coast is mentioned bility is obtained by combining these partial scores (see Tinti as occurred at 08:00 in the morning. The intensity attributed et al., 2004, for details on assigning the ITC parameters). to this tsunami is 3 (Sieberg-Ambraseys scale) and the re- In this research, the result of the scrutiny of new bibli- liability is 2. The estimated Papadopoulos-Imamura scale ographical sources led us to re-appraise the parameters of intensity, hereafter denoted by P-I intensity, is III. some tsunamis that were already included in the ITC. As an example, the 5 February event, that in the catalogue was clas- 3.2 5 February 1783 – time 12:00 sified with Ambraseys intensity 3, has been re-evaluated and now classified with intensity 4. The 5 February tsunami event was at first underestimated and During the 1783–1785 seismic crisis a total of ten considered only a minor event, causing no serious damage. tsunamis occurred all within a time span of about one year On the contrary, the analysis of the sources revealed that in from February 1783 to January 1784: seven were small some localities the tsunami effects were quite strong. The events, two were strong, and one was disastrous, causing Sicily coast from Messina to Torre del Faro (about 11 miles) more than 1500 victims in the small Calabrian village of and in Calabria from Cenidio to Scilla (about 7 miles) was Scilla. In the following, a description of each event is pro- heavily affected by the tsunami (Fig. 4). Historical sources vided, with the specification of the tsunami intensity, relia- refer that generally the sea first receded and then inundated bility and cause. the shore. Withdrawals and inundations repeated at least three times at intervals of about 10–15 min. 3.1 5 February 1783 – time 08:00 At Joppolo a considerable sea withdrawal was observed, that left the sea bottom dry (Sarconi, 1784), and about 8 min This is one of the three new tsunamis discovered during this after the shock in the sea between Nicotera and Gioia Tauro study. According to some sources, at about 08:00 in the two fishermen observed two waves, one travelling toward morning (four hours before the first main shock) a weak Capo Vaticano and the other toward Gioia del Tirreno (Gal- earthquake was felt in Calabria (Fig. 3). On the Ionian side imi, 1783; De Lorenzo, 1877). This latter was as high as a of Calabria, at Capo Rizzuto a sudden sea inundation was ob- big vessel, but caused no damage (Minasi, 1785). served (De Lorenzo, 1877), while at Cutro, some hours be- At Nicotera the sea withdrew boiling and then turned the fore the big earthquake, the sea withdrew and then came back fishing boats upside-down (Vivenzio, 1788). At Gioia del

Nat. Hazards Earth Syst. Sci., 6, 1053–1060, 2006 www.nat-hazards-earth-syst-sci.net/6/1053/2006/ L. Graziani et al.: Revision of the 1783 Calabrian tsunamis 1057

1 Princes House (R=6.2) 2 S. Spirito Church (I=40; R=8) 3 Gornelles fountain (R=8.3) 4 S. Nicola Church (I=100; R>5) 5 Livorno stream (I=200) 6 S. Maria delle Grazie Church (I=90) 1 (R=5.2) 7 2 Chianalea Marina dellOliveto(R=3.5) 8 3 8 4

Marina Grande6 5 7

Fig. 5. Map of Scilla village: the area ﬂooded by the 6 February 1783 tsunami is depicted in grey. Numbers indicate spots for which inundation and/or run-up values have been inferred by the bibliographical sources. Letter “I” indicates the inundation value (in meters), while “R” indicates the run-up value (in meters).

Tirreno the sea inundated the shore without any damage (Mi- (Galimi, 1783). At Roccella, three big waves were observed nasi, 1785). At Scilla the sea withdrew for more than 10 me- about 7 min after the shock. The sea reached the monastery ters leaving the sea bottom dry and a lot of fish on the beach; of the Padri Paolotti, carrying many boats inland (Minasi, then suddenly the water came back surpassing its usual limit 1785; De Lorenzo, 1877). and flooding the coast (Sarconi, 1784). This detail is also It is seen that the tsunami affected a very vast area includ- reported by Minasi (1785, 1971) who tells that at Chianalea ing the southern part of the Tyrrhenian and Ionian coasts of beach (Scilla), the sea withdrew by about 10 m, leaving many Calabria as well as the Messina Straits coast, causing dam- fish aground, and coming back it surpassed the usual shore- age to structures and to boats in some places. As a result of line by about 1.5 m. Minasi (1970) is the only one who fur- this study the tsunami intensity has been re-evaluated, pass- ther mentioned a secondary tsunami occurrence, reporting ing from 3 to 4 (Sieberg-Ambraseys scale), with reliability 4. that at San Gregorio, close to Scilla, a piece of rock fell into We attribute PI intensity VI. the sea generating sea waves. At Messina the sea water got over the quays in the harbour, 3.3 6 February 1783 – time 00:20 inundating and destroying them; the Teatro Marittimo and nearby buildings along the shore were violently inundated Very likely the earthquake is not directly responsible for the (De Leone, 1783; Sarconi, 1784; Hamilton, 1783). The sea tsunami generation and the tsunami was due to a huge rock- remained agitated as long as three days after the main shock. fall that was induced by the earthquake: in fact, a big portion At Torre del Faro (close to Messina), a first sea retreat was of the Mount Campalla` (that is located on the south-western followed by an inundation (Sarconi, 1784). During the with- side of the Scilla beach) collapsed into the sea about half drawal many boats and some fishermen were dragged and an hour after the shock. The front of the failing mass was fish were found on the shore (Anonymous Peloritano, 1783; about 450 m long and penetrated the sea at most by 100 m ac- De Leone, 1783). The water of the two marshes, called Pan- cording to Minasi (1785). From another source (De Lorenzo, tano Piccolo and Pantano Grande, also receded (Vivenzio, 1895) we learn that the front was about 2 km long, with de- 1788). On coming back, the sea partly destroyed the light- tachment occurring at a distance ranging from 500 to 2 km house and penetrated as far as the two marshes, inundating onshore, and that the mass penetrated the sea by a distance vineyards. of 110 m. At Catona the sea flooded the coast for a short stretch (Sar- At Scilla most people, frightened by the sequence of earth- coni, 1784). At Reggio Calabria the sea surpassed the beach quakes of 5 February, escaped to the open beach but dur- arriving at the coastal road (Sarconi, 1784), inundating some ing the night they were surprised by the unexpected waves town districts (De Leone, 1783). In the stretch of coast from (Hamilton, 1783; Sarconi, 1784; Minasi, 1785; Vivenzio, Bianco to Roccella Ionica (about 35 km long) on the Ionian 1788; De Leone, 1783, Torcia, 1783; De Lorenzo, 1877). side of Calabria, the beach was flooded by about one mile A few seconds after the rockfall, the sea violently inundated www.nat-hazards-earth-syst-sci.net/6/1053/2006/ Nat. Hazards Earth Syst. Sci., 6, 1053–1060, 2006 1058 L. Graziani et al.: Revision of the 1783 Calabrian tsunamis

39.200 N 39.200 N 15.200 E 15.200 E CALABRIA CALABRIA N N

S. EUFEMIA GULF S. EUFEMIA GULF Roccella Ionica SQUILLACE SQUILLACE GULF GULF Tropea Bivona

Nicotera Stilo

Torre del Faro Torre del Bagnara Faro Bagnara Scilla Scilla Messina Cannitello Messina Catona Reggio Calabria 37.900 N 37.900 N 17.400 E 17.400 E

Fig. 6. Epicentre of the 6 February 1783 earthquake. Localities hit Fig. 7. Localities affected by the tsunamis of 7 February 1783; of by the generated tsunami are reported. 1 March 1783; of 28 March 1783; of 7 January 1784; of 9 January 1784; and of 19 January 1784. the district of Marina Grande, and the water reached the roof of the buildings (Fig. 5). Three big flooding waves Small houses, people and animals were carried seaward (Au- were observed. The water climbed the Livorno stream up gusti, 1783; Torcia, 1783; Gallo, 1784). At Punta del Pezzo, to 200 m from the sea shore, then reached the beaches called close to Cannitello, and in the surrounding areas, sea cov- Chianalea and Marina dell’Oliveto, that are located on the ered the beach by one and a half mile, leaving sand on the other side (north-east) of the Scilla promontory. At marina ground (Sarconi, 1784). At Reggio Calabria the sea inun- Grande, the Santo Spirito Church was destroyed, and the dated the shore carrying a lot of heavy material (Torcia, 1783; church of Santa Maria delle Grazie (at a distance of about De Leone, 1783). 90 m from the shoreline) was inundated as well as the San Nicola Church. A woman was entangled in a mulberry tree The analysis of the sources put in evidence some new at a height of 8.3 m near the Gornelle fountain (Anonymous data about the tsunami, and confirmed its catastrophic im- Peloritano, 1783; De Lorenzo, 1877). The total number of pact. The event gets the highest reliability and intensity val- victims due to the tsunami exceeded 1500 (Galimi, 1783; ues, exactly as in the ITC, i.e. respectively 4 and 6 (Sieberg- Mercalli, 1906). In Fig. 5 the map of the places affected by Ambraseys scale). We attribute PI intensity X mostly on the the waves is portrayed, with penetration distances and run-up basis of the high toll of human lives in Scilla. values reported. According to Sarconi (1784), in the beach of Marina Grande the observed run-up heights range between 6–9 m, while according to Minasi (1785) the wave reached 3.4 7 February 1783 – time 13:10 the maximum height of 16 m. The tsunami was seen even far from its source (Fig. 6). At Nicotera the sea withdrew and then inundated the shore In the CPTI2 (2004) catalogue an earthquake of equivalent carrying some fishing boats (De Leone, 1783) and also the magnitude Me=6.6 (I=X–XI MCS) is reported as occurred coastal villages of Bagnara and Cannitello were affected inland, about 200 km far from the coast. In the literature, by the inundation (Minasi, 1785; De Lorenzo, 1877). At only one piece of information has been found concerning a Messina the sea was seen to rise and noisily overflowing the tsunami event at Stilo (Fig. 7), along the Ionian Calabrian coast; waves were also quite relevant at the headlight. Minasi coast, when the sea rose with no flooding (Vivenzio, 1788). (1785) reports a sea level rising of about 2 m. At Torre del The historical source mentioning the phenomenon is con- Faro the coast was flooded for about 600 paces. Some boats temporary and it was proven to be very reliable in all cases and 26 victims were carried seaward (Sarconi, 1784; Tor- where cross-checking was possible, and this is reason why cia, 1783; Vivenzio, 1788) and the run-up height was about the tsunami reliability value has been slightly changed, pass- 6 m. At Peloro, near Torre del Faro, a long stretch of fields ing from 1 to 2. The Sieberg-Ambraseys intensity is 2, was flooded, close to the small lake called Pantano Piccolo. whereas the PI intensity is III.

Nat. Hazards Earth Syst. Sci., 6, 1053–1060, 2006 www.nat-hazards-earth-syst-sci.net/6/1053/2006/ L. Graziani et al.: Revision of the 1783 Calabrian tsunamis 1059

3.5 1 March 1783 – time 01:40 only a single place, though with destructive power. Reliabil- ity is 4. Assuming that the ruined houses were light struc- On 1 March 1783 the CPTI2 (2004) catalogue reports a shock tures, the Sieberg- Ambraseys intensity is 3, whereas the in- (equivalent magnitude Me=5.9, I=IX–X MCS) that occurred tensity according to the Papadopoulos and Imamura scale is in land about 100 km far from the Tyrrhenian coast, to the VI. north of the previous earthquakes areas. A contemporary source refers that along the shore at Tropea (Fig. 7) after 3.9 9 January 1784 – time 02:00 the shock a considerable sea flooding was observed (Galimi, 1783) . The available descriptions allow us to attribute to this This is the new identified tsunami event. The available seis- tsunami intensity 2 (Sieberg-Ambraseys scale) and reliability mic catalogues do not report any earthquake in that day. Ac- 3. The PI intensity is III. cording to Vivenzio (1788), at Bivona (Fig. 7) during the night a strong submarine earthquake occurred and a ship in 3.6 24 March 1783 – time 12:00 the harbour broke its mooring and ran aground. The available descriptions allow us to attribute intensity 3 (Sieberg- This is the second new event identified during our study. The Ambraseys scale), PI intensity IV and reliability 2. available seismic catalogues do not report any earthquake oc- 3.10 19 January 1784 curred in that day. During this study we found two sources mentioning a tsunami that occurred on 24 March due to a A light aftershock (Me=4.1, I=IV; ING, 1991) occurred near rockfall of a big portion of the Mount Campalla,` a little east- Messina. An unusual sea agitation was observed between ward to the site that was affected by the 6 February collapse. Torre del Faro and Scilla (Fig. 7). At Faro many banks were There is uncertainty about the year of occurrence: accord- broken. At Fossa and Catona many fields were flooded with ing to one source (Minasi, G., 1971) the rockfall occurred in damage and a lot of trees were overturned (Vivenzio, 1788). 1783 while the other one (Minasi, A., 1970) reports 1790 as Reliability is 4. Putting emphasis on the bank damage at the year of occurrence. Both authors recount that at about Faro, the Sieberg-Ambraseys tsunami intensity is 4. The in- 12:00 a.m. a big portion of Mount Campalla` fell into the sea tensity according to the Papadopoulos and Imamura scale is and that a small boat, that was in the vicinity, was capsized VI. and one man died. It was certainly a local event, that appar- ently did not cause any damage along the shore. Reliability 4 can be attributed to this case. Attributing intensity is quite 4 Conclusions hard, since neither of the two scales (Sieberg-Ambraseys and Papadopoulos-Imamura) seem to fit the case of waves caus- An accurate revision of the 1783–1785 seismic crisis has ing damage to boats offshore, but no damage at the coast. If been carried out, with the main purpose of understanding attention is put on the aspect of boat damaging, the Sieberg- better the tsunami events associated with this crisis. All the Ambraseys intensity can be estimated 3, while the PI inten- events have been evaluated according to the criteria adopted sity can be evaluated IV. for the compilation of the ITC with the aim of updating the catalogue. In addition the Papadopoulos-Imamura tsunami 3.7 28 March 1783 – time 18:55 intensity was evaluated. During this study the thorough analysis of the bibliograph- A very strong shock (Me=6.9, I=X–XI; CPTI2, 2004) caused ical sources permitted us to add more data to the tsunamis severe damage in the villages located between the S.Eufemia already included in the catalogue. For some events the new and the Squillace gulfs, both in the Tyrrhenian and the Ionian information implied a consequent change of the ITC reliabil- coasts (Fig. 7). ity and/or intensity values. At Bagnara, a sea inundation occurred just after the shock Three new tsunami events have been discovered, two (Galimi, 1783). No additional information was found about with intensity 3 and one with intensity 2 on the Sieberg- this event and, therefore, reliability and intensity values re- Ambraseys scale. mained the same, respectively 3 and 2 (Sieberg-Ambraseys scale). The attributed PI intensity value is III. Edited by: E. Pelinovsky Reviewed by: Y. Altinok and G. Papadopoulos 3.8 7 January 1784

A light aftershock (Me=4.1, I=VI; ING, 1991) took place with epicentre near Roccella Ionica. Here the sea ﬂooded most of ﬁelds, causing damage and destroying some small houses (Vivenzio, 1788). No additional information has been found about this event that was very local since it affected www.nat-hazards-earth-syst-sci.net/6/1053/2006/ Nat. Hazards Earth Syst. Sci., 6, 1053–1060, 2006 1060 L. Graziani et al.: Revision of the 1783 Calabrian tsunamis

References Jenny, S., Goes, S., Giardini, D., and Kahle, H. G.: Seismic potential of Southern Italy, Tectonophysics, 415, 81–101, 2006. Ambraseys, N. N.: Data for investigation of seismic sea waves in Mecalli, G.: Alcuni risultati ottenuti dallo studio del terremoto cal- the Eastern Mediterranean, Bull. Seism. Soc. Am., 52, 895–913, abrese dell’ 8 settembre 1905, Atti dell’ Accademia Pontoniana 1962. di Napoli, 36, p. 1–9, 1906. Anonymous Peloritano (Ignazio Paterno’ castello Principe di Bis- Minasi, G.: Continuazione ed appendice sopra i tremuoti descritti cari): Descrizione del Terribile terremoto de’ 5 febbraio 1783 nella relazione colla data di Scilla de 30 settembre 1783, con che afflisse la Sicilia, distrusse Messina e gran parte della Cal- altro che accadde in progresso, Messina, 1785. abria, Napoli, 1783. Minasi, A.: La specola del filosofo, Natura e sorti nelle incisioni di Augusti, M.: Dei terremoti di Messina e di Calabria dell’anno 1783, Antonio Minasi, Ilario Principe, “Brutium”, 1970. memorie e riflessioni, Bologna, 1783. Minasi, G.: Notizie storiche della citta` di Scilla, Edizioni Parallelo Baratta, M.: I Terremoti d’Italia, Arnoldo Forni, Torino, 1901. 38, Reggio Calabria, 1971. Boschi, E., Guidoboni, E., Ferrari, G., Mariotti, D., Valensise, G., Papadopoulos, G. A. and Imamura, F.: Proposal for a new tsunami and Gasperini, P.: Catalogue of strong Italian earthquakes from intensity scale. Proc. Internat.Tsunami Conference, Seattle 7–9 461 b.C. to 1997, Ann. Geofis., 43 609–868, (also available on August 2001, p. 569–577, 2001. CD Rom), 2000. Monaco, C., Tortorici, L., Nicolich, R., Cernobori, L., and Costa, CPTI2 – Gruppo di lavoro CPTI, 2004: Catalogo Parametrico dei M.: From collisional to rifted basins: an example from southern Terremoti Italiani, Versione 2004, INGV, Bologna, 2004. Calabria arc (Italy), Tectonophysics, 266, 233–249, 1996. D’Agostino, N. and Selvaggi, G.: Crustal motion along the Eurasia- Piatanesi, A., Tinti, S., and Gavagni, I.: The slip distribution of the Nubia plate boundary in the Calabrian Arc and Sicily and ac- 1992 Nicaragua earthquake from tsunami run-up data, Geophys. tive extension in the Messina Straits from GPS measurements, Res. Lett., 23, 37–40, 1996. J.Geophys. Res., 109, B11402, doi:10.1029/2004JB002998, Sarconi, M.: Istoria de’ fenomeni del tremuoto avvenuto nelle Cal- 2004. abrie e nel Valdemone nell’anno 1783, Napoli, 1784. De Leone, A.: Giornale e notizie de’ tremuoti accaduti l’anno 1783 Tinti, S. and Maramai, A.: Catalogue of tsunamis generated in Italy nella provincia di Catanzaro, Napoli, 1783. and in Coteˆ d’Azur, France: a step towards a unified catalogue of De Lorenzo, A.: Memorie da servire alla storia sacra e civile di tsunamis in Europe, Annali di Geofisica, 39, 1253–1299, 1996. Reggio e delle Calabrie, Cronache e Documenti inediti o rari, Tinti, S., Maramai, A., and Graziani, L.: The New Catalogue of vol. I, Reggio Calabria, 1877. Italian Tsunamis, Nat. Hazards, 33, 439–465, 2004. De Lorenzo, A.: Un secondo manipolo di monografie e memorie Tinti, S. and Piatanesi, A.: Numerical simulations of the tsunami in- reggine e calabresi, Siena, 1895. duced by the 1627 earthquake affecting Gargano, Southern Italy, Dolomieu, D.: Memoire´ sur le Tremblement de Terre de la Calabre J. Geodynamics, 21, 141–160, 1996a. Pendant l’Annee´ 1783, Fulgoni, Rome, 1784. Tinti, S. and Piatanesi, A.: Finite-element simulations of the 5 Galimi, P.: Lettera di Procopio Galimi al signor D.Giuseppe Vairo February 1783 Calabrian Tsunami, Phys. Chem. Earth, 21(12), sui tremuoti di Calabria dell’anno 1783. 39–43, 1996b. Gallo, A.: Lettera istorico-fisica de’ terremoti accaduti in Messina Torcia, M.: Tremuoto avvenuto nella Calabria e a Messina all 5 nel mese di febbraio di quest’anno 1783, Messina, 8 marzo 1783, febbrajo 1783 descritto da Michele Torcia – Archiviario di S. M. Bologna, 1784. Siciliana e Membro della Accademia Regia, Napoli, 1783. Hamilton, W.: Relazione dell’ultimo terremoto delle Calabrie e Tortorici, G., Bianca, M., De Guidi, G., Monaco, C., and Tortorici, delle Sicilia, inviata da Hamilton, S. E. G. e tradotta dal dottor L.: Fault activity and marine terracing in the Capo Vaticano area Sella, G., Firenze, 1783. (southern Calabria) during the Middle-Late Quaternary, Quat. ING (Istituto Nazionale di Geofisica): Catalogo dei terremoti ital- Int., 101–102, 269–278, 2003. iani dal 1450 a.C. al 1990, Internal file, Roma, 1991. Vivenzio, G.: Istoria de’ tremuoti avvenuti nella provincia di Cal- Jacques, E., Monaco, C., Tapponier, P., Tortrici, L., and Winter, abria ulteriore e nella citta` di Messina nell’anno 1783, Napoli, T.: Faulting and Earthquake triggering during the 1783 Calabria 1788. seismic sequence, Geophys. J. Int., 147, 499–516, 2001.

Nat. Hazards Earth Syst. Sci., 6, 1053–1060, 2006 www.nat-hazards-earth-syst-sci.net/6/1053/2006/